Skip to content
  • There are no suggestions because the search field is empty.

BLAZE Integrations | MCP Usage Limits and Credits

BLAZE Model Context Protocol (MCP) usage is managed through account-level allowances to ensure smooth performance across your organization. Understanding how usage is measured helps you manage your team's activity effectively.

How Usage is Measured

Usage for BLAZE MCP is measured in tool calls. A tool call occurs every time your AI assistant requests information from BLAZE or performs an action in your account.

  • Multiple Calls per Request: A single prompt or complex question may trigger several tool calls. For example, asking "Compare this week's sales across my three stores" may require a separate tool call for each individual store location to pull the necessary data.

Organization-Wide Limits

Usage limits apply to your entire organization, rather than individual users.

NOTE: All team members connecting to BLAZE MCP draw from the same shared account allowance.

What Happens When Limits Are Reached

When your account reaches its MCP usage limit:

  • Assistant Notification: The AI assistant will notify you that the BLAZE MCP usage limit has been reached.
  • Temporary Pause: New requests and actions sent to BLAZE via MCP will pause until your usage allowance resets.
  • No Impact on Operations: Core BLAZE functionality is unaffected—your point-of-sale (POS) systems, back office, and online ordering stores will continue running as normal.

Next Steps & Managing Usage

If you reach your usage limit, you can:

  1. Wait for the Reset: Your usage allowance refreshes automatically at the end of your billing cycle or reset period.
  2. Upgrade Your Plan: Contact your BLAZE representative to upgrade your account plan for a higher tool call allowance.

Tips to Reduce Tool Call Usage

  • Be Specific in Prompts: Specify exact store locations, precise date ranges, and relevant product categories or SKUs.
  • Keep Conversations Focused: Continue related inquiries within the same chat thread so the AI assistant can reference previously fetched data rather than initiating new tool calls.
  • Request Summaries: Ask for high-level summaries instead of complete data lists when you only need an overview.